Pre-launch beta — Pro plan free for the first 100 signups. 0 claimed 100 left Claim →
VISION AI · SCREENSHOT MAKER

Frame-by-Frame Video Screenshot Maker

Extract screenshots every second from YouTube, TikTok, Instagram, Facebook, or uploaded video files. Download ZIP archives and export sequence-ready frames for Vision AI models.

Frame preview
⬇️ Download
0 Selected Click cards or checkboxes
⚡

Developer API & Model Context Protocol (MCP)

Extract frame-by-frame sequences directly inside AI agents (Claude, Cursor) or via REST API.

Add ConvertFleet MCP to your Cursor (~/.cursor/mcp.json) or Claude Desktop (claude_desktop_config.json):
{
  "mcpServers": {
    "convertfleet": {
      "command": "npx",
      "args": ["-y", "convertfleet-mcp@latest"],
      "env": {
        "CONVERTFLEET_API_KEY": "YOUR_API_KEY"
      }
    }
  }
}

Last updated: June 2026

BUILT FOR VISION AI

Designed from the ground up for multimodal AI models.

EACH-SECOND SCREENSHOTS

Precise millisecond timing

Capture exactly 1 frame every second (or 2 fps, 0.5 fps) with synchronized timestamp metadata. Perfect for tracking fast-moving video events and transitions.

ALL SOCIAL NETWORKS

YouTube, TikTok, Reels & FB

Input any public link from YouTube (Shorts & videos), TikTok, Instagram Reels, Facebook, Snapchat, or Twitter. No screen recording or browser extension needed.

DIRECT VIDEO UPLOADER

Upload your own files

Process local recordings, screen captures, and video clips (MP4, MOV, WebM, MKV) up to 250 MB with lightning-fast hardware-accelerated FFmpeg frame extraction.

VISION AI READY

Gemini, GPT-4o & Claude 3.5

Export image sequences with ready-to-use prompt templates for Google Gemini, OpenAI GPT-4o, and Claude. Easily analyze on-screen text, scenes, and object detection.

FAQ

Frequently Asked Questions

How do I extract a screenshot for every second of a video?

Paste any video link (YouTube, TikTok, Instagram, Facebook) or upload a video file, select 'Every 1 second' (or your preferred interval), and click 'Extract Frames'. The engine generates crisp, timestamped image files and a single ZIP package.

How do I use these frames with Vision AI models?

Multimodal models (like Google Gemini 1.5/2.0, OpenAI GPT-4o, and Claude 3.5 Sonnet) can analyze sequential video frames to understand actions, scene changes, on-screen text, and object movements. ConvertFleet provides a 1-click 'Copy Vision AI Prompt' button and direct frame URLs formatted specifically for multimodal agents.

Which social media platforms are supported?

YouTube (regular videos, Shorts), TikTok, Instagram (Reels, posts), Facebook (videos, reels, watch), Snapchat (Spotlight), X (Twitter), Vimeo, and direct video URLs.

Can I upload my own video files directly?

Yes! Switch to the 'Upload Video File' tab to process MP4, MOV, WebM, MKV, or AVI files up to 250 MB directly from your device.

Can I customize the frame rate and image resolution?

Yes. Choose intervals from every 0.5s (2 frames per second) up to every 10s, set safety frame caps (up to 300+ frames), scale images for optimal Vision AI token usage (1024px, 720p, 512px), or burn timestamp badges into the images.